areas, only controlled by a Lagrange parameter Γ 0 . They found that the fluctuation
tension Γ fluc displays a complex behavior and neither agrees with Γ 0 nor with
Γ frame . Schmid [162] has considered an arguably more realistic situation where
“lipids” have fixed area and either a fixed frame tension is applied or the “projected
area” is kept fixed. These simulations reproduce the difference between Γ 0 and
Γ frame and indicate with high accuracy that the fluctuation tension is given by
Γ fluc ¼ Γ frame . Farago [174] confirmed these findings in simulations at fixed
projected area. Furthermore, he carried out reference simulations of a hypothetical
membrane model that lacks gauge invariance and found that, in this case, the
fluctuation tension deviates from the frame tension. These studies support the
validity of the picture originally put forward by Cai et al. [165]: For rotationally
invariant membranes with fixed area per lipid, the fluctuation tension is given by the
frame tension.
2.5 Membrane Heterogeneity and Lipid Rafts
In the late 1990s, several scientists put forward the suggestion that biomembranes
might not be laterally homogeneous but instead contain nanoscopic domains – soon
called “lipid rafts” – which differ in their lipid composition and are important for
numerous membrane-associated biological processes [177–181]. This idea quickly
replaced the prevailing fluid mosaic model [182], according to which the lipid
bilayer merely constitutes a two-dimensional passive solvent that carries membrane
proteins. It created huge excitement due to many obvious biological implications
and possibilities; at the same time it was controversial, for instance because it took
time to converge on a universally accepted definition of what a raft is [82,
183–185].
According to the lipid raft concept, biomembranes are filled with locally phaseseparated, cholesterol-rich, nanoscale “raft” domains, which contribute to membrane heterogeneity and play an important role in organizing the membrane proteins. Two aspects of this hypothesis are well-established: First, biological
membranes are laterally heterogeneous, and heterogeneity is important for the
function of membrane proteins, e.g., in signaling [186]. Second, multicomponent
lipid bilayers phase separate in certain parameter regions into a “liquid disordered”
(ld) and a “liquid ordered” (lo) phase [187, 188]. The hypothetical “raft state” is not
phase-separated, but rather a globally homogeneous state filled with nanodomains
of sizes between 10 and 100 nm. The raft concept is supported by experimental
findings, e.g., on the mobility of certain membrane proteins [189]. It has been
questioned mainly due to a lack of direct evidence. Rafts are too small to be
visualized in vivo by microscopy. Moreover, it was not clear from a theoretical
point of view why nanoscale rafts should be stable with respect to macrophase
separation. To explain this, it was proposed that rafts might be nonequilibrium
structures [190] and that rafts might be stabilized by the cytoplasm [191] or
by special line-active lipids [192–194]. Alternatively, it was argued that rafts
252
M. Deserno et al.
tension Γ fluc displays a complex behavior and neither agrees with Γ 0 nor with
Γ frame . Schmid [162] has considered an arguably more realistic situation where
“lipids” have fixed area and either a fixed frame tension is applied or the “projected
area” is kept fixed. These simulations reproduce the difference between Γ 0 and
Γ frame and indicate with high accuracy that the fluctuation tension is given by
Γ fluc ¼ Γ frame . Farago [174] confirmed these findings in simulations at fixed
projected area. Furthermore, he carried out reference simulations of a hypothetical
membrane model that lacks gauge invariance and found that, in this case, the
fluctuation tension deviates from the frame tension. These studies support the
validity of the picture originally put forward by Cai et al. [165]: For rotationally
invariant membranes with fixed area per lipid, the fluctuation tension is given by the
frame tension.
2.5 Membrane Heterogeneity and Lipid Rafts
In the late 1990s, several scientists put forward the suggestion that biomembranes
might not be laterally homogeneous but instead contain nanoscopic domains – soon
called “lipid rafts” – which differ in their lipid composition and are important for
numerous membrane-associated biological processes [177–181]. This idea quickly
replaced the prevailing fluid mosaic model [182], according to which the lipid
bilayer merely constitutes a two-dimensional passive solvent that carries membrane
proteins. It created huge excitement due to many obvious biological implications
and possibilities; at the same time it was controversial, for instance because it took
time to converge on a universally accepted definition of what a raft is [82,
183–185].
According to the lipid raft concept, biomembranes are filled with locally phaseseparated, cholesterol-rich, nanoscale “raft” domains, which contribute to membrane heterogeneity and play an important role in organizing the membrane proteins. Two aspects of this hypothesis are well-established: First, biological
membranes are laterally heterogeneous, and heterogeneity is important for the
function of membrane proteins, e.g., in signaling [186]. Second, multicomponent
lipid bilayers phase separate in certain parameter regions into a “liquid disordered”
(ld) and a “liquid ordered” (lo) phase [187, 188]. The hypothetical “raft state” is not
phase-separated, but rather a globally homogeneous state filled with nanodomains
of sizes between 10 and 100 nm. The raft concept is supported by experimental
findings, e.g., on the mobility of certain membrane proteins [189]. It has been
questioned mainly due to a lack of direct evidence. Rafts are too small to be
visualized in vivo by microscopy. Moreover, it was not clear from a theoretical
point of view why nanoscale rafts should be stable with respect to macrophase
separation. To explain this, it was proposed that rafts might be nonequilibrium
structures [190] and that rafts might be stabilized by the cytoplasm [191] or
by special line-active lipids [192–194]. Alternatively, it was argued that rafts
252
M. Deserno et al.